[widgitmaster]

Wigit,
Just looked at a previous picture of your machine, looks like you do have a smooth stepper or some other pulse engine plugged into your breakout card. That unit looks like it has a USB plug, so get a high quality USB cable with the Ferrite cord built in and see if all your issues don't disappear.

Russ

Yes, I have a SmoothStepper! I was told it's easier to set the machine up in Mach3 using the Parallel Port 1st, then add the SmoothStepper!
Thanks for all the input too!

Widgit

[kuroguy]

You can pick up a 60Mhz scope on ebay for less than a hundred shipped. Well worth the investment.

[CNCMAN172]

Widget,
You do such nice work it almost makes me cry. Your lathe work is just spectacular, oh how I wish I could crank out stuff like you do. You make brass look like solid gold.

Russ

[widgitmaster]

Thanks Russ!

[widgitmaster]

This morning I finished the z-tool setting plates, putting the setscrew in the bottom and two #2-56 screws near the exit holes.

[widgitmaster]

This afternoon, I rearranged all the ground wires in the VFD box. By putting in a seven position grounding bar, found it in Home Depot! Now all the grounds are neatly organized, instead of the cluster on the single grounding post.

[CarveOne]

Widgit, I'll be making a round trip down I-95 to West Palm Beach this weekend. I'll wave at every city I pass by, knowing that you are in one of them somewhere ... Good luck with the grounding issues.

[widgitmaster]

Widgit, I'll be making a round trip down I-95 to West Palm Beach this weekend. I'll wave at every city I pass by, knowing that you are in one of them somewhere ... Good luck with the grounding issues.

Sorry CarveOne I didn't see ya roll on by, I was hiding in a Hoffman box all day today!





I have totally re-wired both boxes, using shielded cables and extra shielding on everything. So far I have made the 3-axis move, made the spindle & pump run. But there is still too much noise for Mach3 to take control! In order to get the machine to cooperate, I have a "DEBOUNCE" of 130. Not good enough, so I just ordered several dozen ferrite cores, and a shorter DB-25 cable.

After dinner, I called my EE friend, and after a lengthy conversation he pointed out a few error I have made in my wiring! Apparently I had used the frame as my 12-VDC negative! That is a big error, and is letting in allot of the noise! So tomorrow I will re-wire the 12-VDC to correct the problem!

While Googling around, I found an interesting & informative page on electronic noises and how they behave!

[TOTALLYRC]

Filed that away for the next time I work on a noise problem.

Mike

[dodger889]

widgit have thought about doing a Faraday cage around the vfd out of metallic window screening. just make sure that you ground it. that will lessen the rf, emi from it.

[widgitmaster]

As for shields Motor to VFD the shield is to be Grounded at Both Ends, All other Shields should be Grounded at (1) End at the source, & the shield tail should be as short as possible, or if not using a tail a copper saddle placed over the shield, the Grounding points need to be free of Paint Etc to get a metal to metal contact

Can someone explain why the VFD cable's shield is to be grounded on both ends?
Widgit


I think I found the answer!

From the Belden website:

"Do I terminate the shield at both ends?

Drive cable shields should always be terminated at the motor and
at the drive. Intermediate termination of the ground or shield is not
recommended. The shield acts as a conductor for common mode current
containment. Failure to terminate the shields and grounds properly could
result in harmful electrical noise or destructive common mode currents
flowing in the ground grid. Intermediate grounding of the shield can cause
the unintended release of shield currents and resulting electrical noise. "

[ger21]

A debounce setting of 130 is actually very low. Most people with noise issues need to set it to well over 1000-2000 to see any effect.

[CNCMAN172]

Wigit, My guess is the e-stop input into the PC is what is holding you up. The most sensitive pins for the parallel port are clearly the input pins. Most common are the homing sensors and e-stop. You can run a quick experiment to determine if this is your issues. First you need to remove the e-stop wire that is feeding the parallel port, this is normally pin-15. Then you need to go into MACH3 and have it simulate e-stop. Your just doing this to verify the root cause of the problem, this is NOT something you want to keep in place. Once you simulate the e-stop then MACH3 will no longer look at that input pin. In fact you could probably leave everything hooked up and just simulate the e-stop and it would probably accomplish the test. If you are able to get into MACH3 and move the machine around after starting up the spindle you have confirmed the issue. I have also seen the homing sensors fire from EMI noise so during homing the axis would just stop prematurely. My guess is as soon as your fire up the spindle motor you instantly see the e-stop signal in MACH3 firing and you can't get it to reset. Just run this little test and you will know you are on the right path.

In most commercial machines they handle the low voltage logic at 24V via PLC controllers. You can do the same thing by using a relay that is operated at 24VDC and it's contacts would provide the ground to the parallel port, which makes the signal much more immune to noise because it takes 24VDC to activate versus TTL level signal under 5VDC. Since you are planning on moving to the smooth stepper you probably don't want to take this step because you will no longer be using the parallel port.

Anyway give this simple test a try and you might find you can move the machine around.

Russ

[mactec54]

CNCMAN172

What you are suggesting is good, but it does not solve anything, First Wigit has to wire the VFD correctly & then everything will work as it should, A VFD is a noise generator

So you have to solve the EMI/EMF coming from the VFD by wiring it correctly, & then, look at other things if there are still problems, most limit/homing switches like what wigit has used will pick up any Noise, then Trigger a False E-Stop

So he has to fix the source of the Noise, (VFD) before looking at anything else

[widgitmaster]

A little progress today, as I have corrected the 12-VDC circuits. The VFD cable is grounded on both ends, and ALL cables entering both the Hoffman box's are shielded & grounded. Now I do have a problem on Pin #15, which is interfering with Mach3's ability to home the machine. So I'll have to wait until all the Ferrite-Cores arrive, and the new DB-25 cable. The problem with Pin #15 is inside the gantry boxes, and not the result of the VFD's noise!
After studying the way I wired the e-stops, I think I may have a solution for it! But right now, I'm too tired & sore to do any more!

Oh yea,
I have the DEBOUNCE set to zero when the VFD is off, and 50 when its running!

[dodger889]

Hey guys just wondering if you where to use a relay( solid state) in the E stop circuit. to control what mach is seeing if that might solve this problem? Here is my thinking about that it could be isolated from the rest of the control box in it's own little enclosure. This should get away from the interference that is happening.

[CNCMAN172]

Dodger889,
You are correct this method works well, and is what I described about as the 24VDC relay. I actually implemented this myself due to the touchy nature and the noise fountains these Chinese spindles makes. LOL... Honestly, thought for the money I have put many hours on my 4KW spindle and cut lots and lots of aluminum and it works way better than my Bosch Wood Router, not question, the noise level is also much lower.
Russ

[widgitmaster]

So your saying I should add another relay between the breakout board and the E-stop buttons, that is not as difficult as it sounds and would decrease the pin #15 E-Stop cable to about 14" instead of 19-feet!
I like it, thank you!
Widgit

[CNCMAN172]

Wigit, with 19 feet of cable running to the estop buttons that is a huge area to pickup noise. Here is what you need to do to make this much more noise immune. The long wires need to carry the higher voltage that will energize a relay, say 12Volts or 24Volts. The contact on the relay is located close to the breakout board and will provide the ground to pin-15. One side of the contact will be grounded the other side ties to pin-15. Most parallel ports pull up the input pins so it is normally high when activated it will get pulled low by the relay. Once you have that all working you will most likely need to do something on home sensors as well. Not sure what you used for homing sensors, switches optical sensors, etc. Send that info and we will provide a solution for that aspect as well. This will address the e-stop.

Russ

[widgitmaster]

Actually, there is an ice cube relay connected to the e-stops already, as each e-stop has two normally closed switch blocks. The 1st relay controls the VFD's contactor!